AdaBoost

AdaBoost constructs an additive classifier by repeatedly fitting a base classifier to exponentially reweighted observations. Misclassified observations receive greater weight at the next iteration, and the final score is a nonnegative weighted sum of the base classifiers.